Objective To explore the application of artificial intelligence technology in individualized depression care for patients with depression, to achieve precise care goal, and to accelerate the rehabilitation of them. Methods Totally, 60 patients with depression were evenly and randomly assigned to a control group and an intervention group in the study. The control group received traditional nursing, while the intervention group received individualized nursing care based on deep learning emotion classification models:firstly, EEG data with labels were captured by electronic devices; secondly, the emotion classification model based on deep learning was used to recognize the real emotion of patients; thirdly, individualized nursing strategy was selected according to the recognized result. Quantitative assessment of depression and nursing satisfaction of patients in both groups during hospitalization were investigated. Results Four weeks into the intervention, the intervention group had significantly lower HAMD scores and SDS scores than the controls （P<0.05 for both）. Eight weeks into the intervention, the intervention group had higher recovery rate, though the difference between the 2 groups was not significant （P>0.05）. Conclusion Individualized depression nursing strategy based on deep learning emotion classification models can significantly relieve the degree of depression and accelerate the recovery of patients.